The 2018 and 2019 Northwest Pacific Typhoon seasons resulted in Japan experiencing some of the costliest typhoon-induced insured losses on record. Current estimates for Jebi, Hagibis and Faxai rank them the first, second and fourth largest storm-related catastrophic losses in Japan when compared with estimates from the General Insurance Association of Japan. 2018’s Trami was another event which additionally resulted in a multi-billion-dollar payout for insurers.

Although Jebi is the largest loss on record, Japanese insurance companies’ reinsurance programmes worked effectively. Japanese companies have been using a wide variety of risk modelling tools for a long time, boosting their understanding of typhoon risk and its financial impact in Japan. This risk management maturity has led to the current adequate reinsurance limits purchase.

Jebi and Hagibis have shown two unique features.

• Firstly, Jebi is the first major hurricane-equivalent (Category 3 or stronger) to have a direct hit on a large Japanese metropolis (Osaka). It is possible that high winds combined with a landscape of skyscrapers and wide streets caused unforeseen localized effects, increasing the experienced wind gusts relatively compared to the same typhoon passing over less urban areas. This process around skyscrapers is known as the Venturi Effect, when winds flow through a constricted passage resultingly in higher speeds.
• Secondly, Hagibis, which showed much lower intensity wind gusts than Jebi, produced an extraordinary amount of rainfall with some areas accumulating more than 1m of precipitation in 24h. According to reports by Japanese broadcaster NKH on October 15, 2019, Hagibis caused more than 90 levee breaches following unpreceded levels of rainfall and flooding. These however were within expectation of the risk levels set by the government risk mitigation strategy and flood maps.

Vendor modeling companies have struggled to provide consistently reliable post loss estimates, creating uncertainty in the market. In response, Aon developed an alternative tool that combines real time measurements from the Japan Meteorological Agency (JMA) stations with damage curves from Aon’s Impact Forecasting team that have been built with local data. This new tool helps to pinpoint the total number of buildings damaged hours after an event occurrence to develop reliable claims estimates and help insurers reach more informed decisions during challenging times.

Outputs and benchmarking of the claims and loss estimates are also used to test Impact Forecasting’s new Japan typhoon model, taking into account what has been learned (and still learning) from these recent storms (Jebi, Trami, Hagibis, Faxai).

This is an extract from: Weather, Climate & Catastrophe Insight Report: 2019. To access the full report, please click here. 

About the Author

Oriol Gaspa leads the Analytics team for Aon’s Reinsurance Solutions business in Tokyo. He develops the catastrophe management, actuarial and research & development teams to provide risk assessment to insurance clients and their reinsurance markets to enable appropriate risk transfer. He previously focused on model evaluation London and has also developed earthquake and tsunami catastrophe models. Before Aon, Oriol worked at the International Seismological Centre as a seismologist and developer. He holds an MSc. in physics and a Msc. in Seismology from the University of Barcelona.